A Note on the SPICE Method

TL;DR

This paper analyzes the SPICE method, linking it to Lasso techniques, and demonstrates its efficiency and potential for reliable sparse estimation in various scenarios.

Contribution

It establishes the connections between SPICE and standard sparse estimation methods, enhancing understanding of its properties and suggesting improvements.

Findings

SPICE is computationally efficient for Lasso-type estimators.

Connections enable analysis of SPICE's asymptotic properties.

Modifications improve SPICE's performance in challenging scenarios.

Abstract

In this article, we analyze the SPICE method developed in [1], and establish its connections with other standard sparse estimation methods such as the Lasso and the LAD-Lasso. This result positions SPICE as a computationally efficient technique for the calculation of Lasso-type estimators. Conversely, this connection is very useful for establishing the asymptotic properties of SPICE under several problem scenarios and for suggesting suitable modifications in cases where the naive version of SPICE would not work.